Development and validation of a machine learning model to predict the use of renal replacement therapy in 14,374 patients with COVID-19.

PURPOSE: To develop a model to predict the use of renal replacement therapy (RRT) in COVID-19 patients. MATERIALS AND METHODS: Retrospective analysis of multicenter cohort of intensive care unit (ICU) admissions of Brazil involving COVID-19 critically adult patients, requiring ventilatory support, admitted to 126 Brazilian ICUs, from February 2020 to December 2021 (development) and January to May 2022 (validation). No interventions were performed. RESULTS: Eight machine learning models' classifications were evaluated. Models were developed using an 80/20 testing/train split ratio and cross-validation. Thirteen candidate predictors were selected using the Recursive Feature Elimination (RFE) algorithm. Discrimination and calibration were assessed. Temporal validation was performed using data from 2022. Of 14,374 COVID-19 patients with initial respiratory support, 1924 (13%) required RRT. RRT patients were older (65 [53-75] vs. 55 [42-68]), had more comorbidities (Charlson's Comorbidity Index 1.0 [0.00-2.00] vs 0.0 [0.00-1.00]), had higher severity (SAPS-3 median: 61 [51-74] vs 48 [41-58]), and had higher in-hospital mortality (71% vs 22%) compared to non-RRT. Risk factors for RRT, such as Creatinine, Glasgow Coma Scale, Urea, Invasive Mechanical Ventilation, Age, Chronic Kidney Disease, Platelets count, Vasopressors, Noninvasive Ventilation, Hypertension, Diabetes, modified frailty index (mFI) and Gender, were identified. The best discrimination and calibration were found in the Random Forest (AUC [95%CI]: 0.78 [0.75-0.81] and Brier's Score: 0.09 [95%CI: 0.08-0.10]). The final model (Random Forest) showed comparable performance in the temporal validation (AUC [95%CI]: 0.79 [0.75-0.84] and Brier's Score, 0.08 [95%CI: 0.08-0.1]). CONCLUSIONS: An early ML model using easily available clinical and laboratory data accurately predicted the use of RRT in critically ill patients with COVID-19. Our study demonstrates that using ML techniques is feasible to provide early prediction of use of RRT in COVID-19 patients.

A Prospective International Multicenter Study of AKI in the Intensive Care Unit.

BACKGROUND AND OBJECTIVES: AKI is frequent and is associated with poor outcomes. There is limited information on the epidemiology of AKI worldwide. This study compared patients with AKI in emerging and developed countries to determine the association of clinical factors and processes of care with outcomes. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: This prospective observational study was conducted among intensive care unit patients from nine centers in developed countries and five centers in emerging countries. AKI was defined as an increase in creatinine of ≥0.3 mg/dl within 48 hours. RESULTS: Between 2008 and 2012, 6647 patients were screened, of whom 1275 (19.2%) developed AKI. A total of 745 (58% of those with AKI) agreed to participate and had complete data. Patients in developed countries had more sepsis (52.1% versus 38.0%) and higher Acute Physiology and Chronic Health Evaluation (APACHE) scores (mean±SD, 61.1±27.5 versus 51.1±25.2); those from emerging countries had more CKD (54.3% versus 38.3%), GN (6.3% versus 0.9%), and interstitial nephritis (7.0% versus 0.6%) (all P<0.05). Patients from developed countries were less often treated with dialysis (15.5% versus 30.2%; P<0.001) and started dialysis later after AKI diagnosis (2.0 [interquartile range, 0.75-5.0] days versus 0 [interquartile range, 0-5.0] days; P=0.02). Hospital mortality was 22.0%, and 13.3% of survivors were dialysis dependent at discharge. Independent risk factors associated with hospital mortality included older age, residence in an emerging country, use of vasopressors (emerging countries only), dialysis and mechanical ventilation, and higher APACHE score and cumulative fluid balance (developed countries only). A lower probability of renal recovery was associated with residence in an emerging country, higher APACHE score (emerging countries only) and dialysis, while mechanical ventilation was associated with renal recovery (developed countries only). CONCLUSIONS: This study contrasts the clinical features and management of AKI and demonstrates worse outcomes in emerging than in developed countries. Differences in variations in care may explain these findings and should be considered in future trials.

Predialysis hypernatremia is a prognostic marker in acute kidney injury in need of renal replacement therapy.

BACKGROUND AND OBJECTIVES: The present study aimed to evaluate the prognostic impact of predialysis dysnatremia in patients with acute kidney injury requiring renal replacement therapy (RRT). DESIGN, SETTING, PARTICIPANTS, AND MEASUREMENTS: A secondary analysis of a prospective multicenter cohort study was performed. Serum sodium (Na) concentrations were categorized immediately before the first RRT as normonatremia (135≤Na ≤145mEq/L), hyponatremia (mild [130≤Na ≤134mEq/L] or severe [Na ≤129mEq/L]), and hypernatremia (mild [146≤Na ≤155mEq/L] or severe [Na ≥156mEq/L]). Multivariable logistic regression was used to estimate the impact of sodium levels categories on hospital mortality. RESULTS: Dysnatremia occurred in 47.3% of 772 included patients. Hypernatremia was more frequent than hyponatremia (33.7% vs 13.6%, P=.001). Intensive care unit (ICU) and hospital mortality rates were 64.6% and 69%, respectively. Hospital mortality was higher in severe hypernatremia (89.1% [95% confidence interval {CI}, 78.7%-95.8%] vs 64.6% [CI, 59.8%-69.2%], P<.001, in normonatremia). Older patients, clinical admission, number of comorbidities, length of ICU stay before the beginning of RRT, and the number of organ dysfunctions were associated with higher hospital mortality. In multivariate analysis, severe hypernatremia (odds ratio, 2.87; 95% CI, 1.2-6.9), poor chronic heath status, severity of illness, sepsis, and lactate were independently associated with outcome. CONCLUSION: Almost 50% of patients with acute kidney injury in need of RRT in the ICU had mild or severe dysnatremia before dialysis initiation. Hypernatremia was the main sodium disturbance and independently associated with poor outcome in the study population.

The evaluation of sequential platelet counts has prognostic value for acute kidney injury patients requiring dialysis in the intensive care setting.

OBJECTIVE: To evaluate the prognostic value of platelet counts in acute kidney injury patients requiring renal replacement therapy. METHODS: This prospective cohort study was performed in three tertiary-care hospitals. Platelet counts were obtained upon admission to the intensive care unit and during the first week of renal replacement therapy on days 1, 3, 5 and 7. The outcome of interest was the hospital mortality rate. With the aim of minimizing individual variation, we analyzed the relative platelet counts on days 3, 5, 7 and at the point of the largest variation during the first week of renal replacement therapy. Logistic regression analysis was used to test the prognostic value of the platelet counts. RESULTS: The study included 274 patients. The hospital mortality rate was 62%. The survivors had significantly higher platelet counts upon admission to the intensive care unit compared to the non-survivors [175.5×10(3)/mm(3) (108.5-259×10(3)/mm(3)) vs. 148×10(3)/mm(3) (80-141×10(3)/mm(3))] and during the first week of renal replacement therapy. The relative platelet count reductions were significantly associated with a higher hospital mortality rate compared with the platelet count increases (70% vs. 44% at the nadir, respectively). A relative platelet count reduction >60% was significantly associated with a worse outcome (mortality rate=82.6%). Relative platelet count variations and the percentage of reduction were independent risk factors of hospital mortality during the first week of renal replacement therapy. CONCLUSION: Platelet counts upon admission to the intensive care unit and at the beginning of renal replacement therapy as well as sequential platelet count evaluation have prognostic value in acute kidney injury patients requiring renal replacement therapy.

The evaluation of sequential platelet counts has prognostic value for acute kidney injury patients requiring dialysis in the intensive care setting

OBJECTIVE: To evaluate the prognostic value of platelet counts in acute kidney injury patients requiring renal replacement therapy. METHODS: This prospective cohort study was performed in three tertiary-care hospitals. Platelet counts were obtained upon admission to the intensive care unit and during the first week of renal replacement therapy on days 1, 3, 5 and 7. The outcome of interest was the hospital mortality rate. With the aim of minimizing individual variation, we analyzed the relative platelet counts on days 3, 5, 7 and at the point of the largest variation during the first week of renal replacement therapy. Logistic regression analysis was used to test the prognostic value of the platelet counts. RESULTS: The study included 274 patients. The hospital mortality rate was 62%. The survivors had significantly higher platelet counts upon admission to the intensive care unit compared to the non-survivors [175.5×103/mm3 (108.5-259×103/mm3) vs. 148×103/mm3 (80−141×103/mm3)] and during the first week of renal replacement therapy. The relative platelet count reductions were significantly associated with a higher hospital mortality rate compared with the platelet count increases (70% vs. 44% at the nadir, respectively). A relative platelet count reduction >60% was significantly associated with a worse outcome (mortality rate = 82.6%). Relative platelet count variations and the percentage of reduction were independent risk factors of hospital mortality during the first week of renal replacement therapy. CONCLUSION: Platelet counts upon admission to the intensive care unit and at the beginning of renal replacement therapy as well as sequential platelet count evaluation have prognostic value in acute kidney injury patients requiring renal replacement therapy. .

A20-mediated modulation of inflammatory and immune responses in aortic allografts and development of transplant arteriosclerosis.

BACKGROUND: Transplant arteriosclerosis (TA) is the pathognomonic feature of chronic rejection, the primary cause of allograft failure. We have shown that the NF-κB inhibitory protein A20 exerts vasculoprotective effects in endothelial and smooth muscle cells (SMC), and hence is a candidate to prevent TA. We sought direct proof for this hypothesis. METHODS: Fully mismatched, C57BL/6 (H2) into BALB/c (H2), aorta to carotid allografts were preperfused with saline, recombinant A20 adenovirus (rAd.A20) or rAd.ß-galactosidase (ß-gal), implanted, harvested 4 weeks after transplantation, and analyzed by histology, immunohistochemistry, and immunofluorescence staining. We measured indoleamine 2,3-dioxygenase, interleukin-6, and transforming growth factor-ß mRNA and protein levels in nontransduced, and rAd.A20 or rAd.ß-gal-transduced human SMC cultures after cytokine treatment. RESULTS: Vascular overexpression of A20 significantly reduced TA lesions. This correlated with decreased graft inflammation and increased apoptosis of neointimal SMC. Paradoxically, T-cell infiltrates increased in A20-expressing allografts, including the immunoprivileged media, which related to A20 preventing indoleamine 2,3-dioxygenase upregulation in SMC. However, infiltrating T cells were predominantly T-regulatory cells (CD25+/Forkhead Box P3 [FoxP3+]). This agrees with A20 inhibiting interleukin-6 and promoting transforming growth factor-ß production by medial SMC and in SMC cultures exposed to cytokines, which favors differentiation of regulatory over pathogenic T cells. CONCLUSIONS: In summary, A20 prevents immune-mediated remodeling of vascular allografts, therefore reduces TA lesions by affecting apoptotic and inflammatory signals and modifying the local cytokine milieu to promote an immunoregulatory response within the vessel wall. This highlights a novel function for A20 in local immunosurveillance, which added to its vasculoprotective effects, supports its therapeutic promise in TA.

Hepatocyte growth factor preferentially activates the anti-inflammatory arm of NF-κB signaling to induce A20 and protect renal proximal tubular epithelial cells from inflammation.

Inflammation induces the NF-κB dependent protein A20 in human renal proximal tubular epithelial cells (RPTEC), which secondarily contains inflammation by shutting down NF-κB activation. We surmised that inducing A20 without engaging the pro-inflammatory arm of NF-κB could improve outcomes in kidney disease. We showed that hepatocyte growth factor (HGF) increases A20 mRNA and protein levels in RPTEC without causing inflammation. Upregulation of A20 by HGF was NF-κB/RelA dependent as it was abolished by overexpressing IκBα or silencing p65/RelA. Unlike TNFα, HGF caused minimal IκBα and p65/RelA phosphorylation, with moderate IκBα degradation. Upstream, HGF led to robust and sustained AKT activation, which was required for p65 phosphorylation and A20 upregulation. While HGF treatment of RPTEC significantly increased A20 mRNA, it failed to induce NF-κB dependent, pro-inflammatory MCP-1, VCAM-1, and ICAM-1 mRNA. This indicates that HGF preferentially upregulates protective (A20) over pro-inflammatory NF-κB dependent genes. Upregulation of A20 supported the anti-inflammatory effects of HGF in RPTEC. HGF pretreatment significantly attenuated TNFα-mediated increase of ICAM-1, a finding partially reversed by silencing A20. In conclusion, this is the first demonstration that HGF activates an AKT-p65/RelA pathway to preferentially induce A20 but not inflammatory molecules. This could be highly desirable in acute and chronic renal injury where A20-based anti-inflammatory therapies are beneficial.

Outcomes of cancer and non-cancer patients with acute kidney injury and need of renal replacement therapy admitted to general intensive care units.

BACKGROUND: Studies on cancer patients with acute kidney injury (AKI) are restricted to specialized intensive care units (ICUs). The aim of this study was to compare the characteristics and outcomes of cancer and non-cancer patients requiring renal replacement therapy (RRT) for AKI in general ICUs. METHODS: A prospective cohort study was conducted in 14 ICUs from three tertiary care hospitals. A total of 773 (non-cancer 85%; cancer 15%) consecutive patients were included over a 44-month period. Logistic regression was used to identify factors associated with hospital mortality. RESULTS: Continuous RRT was used in 79% patients. The main contributing factors for AKI were sepsis (72%) and ischaemia/shock (66%); AKI was multifactorial in 87% of cancer and in 71% non-cancer patients. Hospital mortality rates were higher in cancer (78%) than in non-cancer patients (68%) (P=0.042). However, in multivariate analyses, older age, medical admission, poor chronic health status, comorbidities, ICU days until the RRT start and number of associated organ dysfunctions were associated with hospital mortality. The diagnosis of cancer was not independently associated with mortality [odds ratio=1.54 (95% confidence interval, 0.88-2.62), P=0.115]. Mortality in cancer patients was mostly dependent on the number of associated organ dysfunctions. Of note, 85% cancer patients recovered renal function at hospital discharge. CONCLUSIONS: In general ICUs, one in six patients requiring RRT has cancer. Despite a relatively higher mortality, the presence of cancer was not independently associated with mortality in the present cohort.

O-glycosylation regulates ubiquitination and degradation of the anti-inflammatory protein A20 to accelerate atherosclerosis in diabetic ApoE-null mice.

BACKGROUND: Accelerated atherosclerosis is the leading cause of morbidity and mortality in diabetic patients. Hyperglycemia is a recognized independent risk factor for heightened atherogenesis in diabetes mellitus (DM). However, our understanding of the mechanisms underlying glucose damage to the vasculature remains incomplete. METHODOLOGY/PRINCIPAL FINDINGS: High glucose and hyperglycemia reduced upregulation of the NF-κB inhibitory and atheroprotective protein A20 in human coronary endothelial (EC) and smooth muscle cell (SMC) cultures challenged with Tumor Necrosis Factor alpha (TNF), aortae of diabetic mice following Lipopolysaccharide (LPS) injection used as an inflammatory insult and in failed vein-grafts of diabetic patients. Decreased vascular expression of A20 did not relate to defective transcription, as A20 mRNA levels were similar or even higher in EC/SMC cultured in high glucose, in vessels of diabetic C57BL/6 and FBV/N mice, and in failed vein grafts of diabetic patients, when compared to controls. Rather, decreased A20 expression correlated with post-translational O-Glucosamine-N-Acetylation (O-GlcNAcylation) and ubiquitination of A20, targeting it for proteasomal degradation. Restoring A20 levels by inhibiting O-GlcNAcylation, blocking proteasome activity, or overexpressing A20, blocked upregulation of the receptor for advanced glycation end-products (RAGE) and phosphorylation of PKCßII, two prime atherogenic signals triggered by high glucose in EC/SMC. A20 gene transfer to the aortic arch of diabetic ApoE null mice that develop accelerated atherosclerosis, attenuated vascular expression of RAGE and phospho-PKCßII, significantly reducing atherosclerosis. CONCLUSIONS: High glucose/hyperglycemia regulate vascular A20 expression via O-GlcNAcylation-dependent ubiquitination and proteasomal degradation. This could be key to the pathogenesis of accelerated atherosclerosis in diabetes.

A20 inhibits post-angioplasty restenosis by blocking macrophage trafficking and decreasing adventitial neovascularization.

OBJECTIVE: Neointimal hyperplasia is an inflammatory and proliferative process that occurs as a result of injury to the vessel wall. We have shown that the homeostatic protein A20 prevents neointimal hyperplasia by affecting endothelial cell (EC) and smooth muscle cell (SMC) responses to injury. In this work, we questioned whether A20 impacts other pathogenic effectors of neointimal hyperplasia including homing of monocyte/macrophages and EC/SMC precursors to the site of vascular injury, vascular endothelial growth factor (VEGF) secretion, and adventitial neovascularization. METHODS AND RESULTS: Carotid balloon angioplasty was performed on rat recipients of a bone marrow transplant from green fluorescent rats. Adenoviral delivery of A20 prevented neointimal hyperplasia and decreased macrophage infiltration. This was associated with decreased ICAM-1 and MCP-1 expression in vitro. Additionally, A20 reduced neovascularization in the adventitia of balloon injured carotid arteries, which correlated with fewer VEGF positive cells. CONCLUSIONS: A20 downregulates adhesion markers, chemokine production, and adventitial angiogenesis, all of which are required for macrophage trafficking to sites of vascular injury. This, in turn, diminishes the inflammatory milieu to prevent neointimal hyperplasia.

SAPS 3 scores at the start of renal replacement therapy predict mortality in critically ill patients with acute kidney injury.

Patients can experience acute kidney injury and require renal replacement therapy at any time during their admission to intensive care units. Prognostic scores have been used to characterize and stratify patients by the severity of acute disease, but scores based on findings during the day of admission may not be reliable surrogate markers of the severity of acute illness in this population. The aim of this study was to evaluate the performance of SAPS 3 and MPM(0)-III scores, determined at the start of renal replacement therapy, in 244 patients admitted to 11 units of three hospitals in Rio de Janeiro, Brazil. Continuous renal replacement therapy was used as first indication in 84% of these patients. Discrimination by area under the receiver operating characteristic curve was significantly better for SAPS 3 than for MPM(0)-III, as was the calibration measured by the Hosmer-Lemeshow goodness-of-fit test. Mortality prediction and calibration approached those eventually found when a customized equation of SAPS 3 for Central and South America was used. After adjusting for other relevant covariates in multivariate analyses, both higher prognostic scores and length of stay in the unit prior to the start of renal replacement therapy were the main predictive factors for hospital mortality. Our study shows that a customized SAPS 3 model was accurate in predicting mortality and seems a promising algorithm to characterize and stratify patients in clinical studies.

Outcomes of critically ill patients with acute kidney injury and end-stage renal disease requiring renal replacement therapy: a case-control study.

BACKGROUND: This study aimed to evaluate and compare the characteristics and outcomes of patients with end-stage renal disease (ESRD) with those of matched controls of patients with acute kidney injury (AKI) requiring renal replacement therapy. METHODS: A case-control study was performed at the intensive care units (ICU) of three tertiary-care hospitals between December 2004 and September 2007. Patients were admitted with life-threatening complications and were matched for age and for severity of illness and organ dysfunctions. Conditional logistic regression was used to identify factors associated with hospital mortality. RESULTS: A total of 54 patients with ESRD and 54 patients with AKI were eligible for the study and were well matched. In general, clinical characteristics were similar. Nonetheless, comorbidities were more frequent in patients with ESRD, and patients with AKI more frequently required mechanical ventilation. ICU (43% versus 20%, P = 0.023) and hospital (50% versus 24%, P = 0.010) mortality rates were higher in patients with AKI. In addition, patients with AKI experienced longer ICU and hospitals stays. The SAPS II score had a regular ability in discriminating survivors and non-survivors, and tended to underestimate mortality in patients with AKI and overestimate in patients with ESRD. When all patients were evaluated, older age [OR = 1.05 (95% CI, 1.01-1.09)], poor chronic health status [OR = 3.90(1.19-12.82)] and number of associated organ failures [OR = 4.44(1.97-10.00)] were the main independent predictors of mortality. After adjusting for those covariates, ESRD was still associated with a lower probability of death [OR = 0.17 (0.06-0.050)]. CONCLUSIONS: ESRD patients with life-threatening complications had significantly better outcome than AKI patients.

Effects of early changes in organ dysfunctions on the outcomes of critically ill patients in need of renal replacement therapy.

INTRODUCTION: Acute kidney injury usually develops in critically ill patients in the context of multiple organ dysfunctions. OBJECTIVE: To evaluate the effect of changes in associated organ dysfunctions over the first three days of renal replacement therapy on the outcomes of patients with acute kidney injury. METHODS: Over a 19-month period, we evaluated 260 patients admitted to the intensive care units of three tertiary-care hospitals who required renal replacement therapy for > 48 h. Organ dysfunctions were evaluated by SOFA score (excluding renal points) on the first (D1) and third (D3) days of renal replacement therapy. Absolute (A-SOFA) and relative (Delta-SOFA) changes in SOFA scores were also calculated. RESULTS: Hospital mortality rate was 75%. Organ dysfunctions worsened (A-SOFA>0) in 53%, remained unchanged (A-SOFA=0) in 17% and improved (A-SOFA<0) in 30% of patients; and mortality was lower in the last group (80% vs. 84% vs. 61%, p=0.003). SOFA on D1 (p<0.001), SOFA on D3 (p<0.001), A-SOFA (p=0.019) and Delta-SOFA (p=0.016) were higher in non-survivors. However, neither A-SOFA nor Delta-SOFA discriminated survivors from non-survivors on an individual basis. Adjusting for other covariates (including SOFA on D1), A-SOFA and Delta-SOFA were associated with increased mortality, and patients in whom SOFA scores worsened or remained unchanged had poorer outcomes. CONCLUSIONS: In addition to baseline values, early changes in SOFA score after the start of renal replacement therapy were associated with hospital mortality. However, no prognostic score should be used as the only parameter to predict individual outcomes.

Desempenho de seis modelos de predição prognóstica em pacientes críticos que receberam suporte renal extracorpóreo / Performance of six prognostic scores in critically ILL patients receiving renal replacement therapy

JUSTIFICATIVA E OBJETIVOS: Não existe consenso sobre qual modelo prognóstico deva ser utilizado em pacientes com disfunção renal aguda (DRA). O objetivo deste estudo foi avaliar o desempenho de seis escores de prognóstico em pacientes que necessitaram de suporte renal. MÉTODO: Coorte prospectiva de pacientes internados nas unidades de terapia intensiva (UTI) de três hospitais terciários que necessitaram de suporte renal por DRA durante 32 meses. Foram excluídos os pacientes crônicos em programa de diálise ou com < 24h de internação na UTI. Os dados das primeiras 24h de UTI foram utilizados no cálculo do SAPS II e do APACHE II, e os dados das primeiras 24h de suporte renal foram utilizados no cálculo dos escores LOD, ODIN, Liaño e Mehta. A discriminação foi avaliada através da área sobre a curva ROC (AUROC) e a calibração através do teste do goodness-of-fit de Hosmer-Lemeshow. A letalidade hospitalar foi o desfecho de interesse. RESULTADOS: Quatrocentos e sesseta e sete pacientes foram incluídos e a letalidade hospitalar foi 75 por cento. Os valores dos escores SAPS II, APACHE II e LOD foram 48,5 ± 11,2, 27,4 ± 6,3, 7 (5-8) pontos, respectivamente. A calibração foi adequada para todos os escores, com exceção do Mehta (p = 0,001). Entretanto, a discriminação foi ruim para todos os modelos, com AUROC variando entre 0,60 para o ODIN e 0,72 para o SAPS II e Mehta. Com exceção do Mehta, todos os modelos subestimaram a letalidade. CONCLUSÕES: Todos os seis modelos estudados foram inadequados na predição prognóstica de pacientes graves com DRA e necessidade de suporte renal.

BACKGROUND AND OBJECTIVES: There is no consensus about prognostic scores for use in patients with acute kidney injury (AKI). The aim of this study was to evaluate the performance of six prognostic scores in predicting hospital mortality in patients with AKI and need for renal replacement therapy (RRT). METHODS: Prospective cohort of patients admitted to the intensive care units (ICU) of three tertiary care hospitals that required RRT for AKI over a 32-month period. Patients with end-stage renal disease and those with ICU stay < 24h were excluded. Data from the first 24h of ICU admission were used to calculate SAPS II and APACHE II scores, and data from the first 24h of RRT were used in the calculation of LOD, ODIN, Liaño and Mehta scores. Discrimination was evaluated using the area under ROC curve (AUROC) and calibration using the Hosmer-Lemeshow goodness-of-fit test. The hospital mortality was the end-point of interest. RESULTS: 467 patients were evaluated. Hospital mortality rate was 75 percent. Mean SAPS II and APACHE II scores were 48.5 ±11.2 and 27.4 ± 6.3 points, and median LOD score was 7 (5-8) points. Except for Mehta score (p = 0.001), calibration was appropriate in all models. However, discrimination was uniformly unsatisfactory; AUROC ranged from 0.60 for ODIN to 0.72 for SAPS II and Mehta scores. In addition, except for Mehta, all models tended to underestimate hospital mortality. CONCLUSIONS: Organ dysfunction, general and renal-specific severity-of-illness scores were inaccurate in predicting outcome in ICU patients in need for RRT.

Performance of six prognostic scores in critically ILL patients receiving renal replacement therapy. / Desempenho de seis modelos de predição prognóstica em pacientes críticos que receberam suporte renal extracorpóreo.

BACKGROUND AND OBJECTIVES: There is no consensus about prognostic scores for use in patients with acute kidney injury (AKI). The aim of this study was to evaluate the performance of six prognostic scores in predicting hospital mortality in patients with AKI and need for renal replacement therapy (RRT). METHODS: Prospective cohort of patients admitted to the intensive care units (ICU) of three tertiary care hospitals that required RRT for AKI over a 32-month period. Patients with end-stage renal disease and those with ICU stay < 24h were excluded. Data from the first 24h of ICU admission were used to calculate SAPS II and APACHE II scores, and data from the first 24h of RRT were used in the calculation of LOD, ODIN, Liaño and Mehta scores. Discrimination was evaluated using the area under ROC curve (AUROC) and calibration using the Hosmer-Lemeshow goodness-of-fit test. The hospital mortality was the end-point of interest. RESULTS: 467 patients were evaluated. Hospital mortality rate was 75%. Mean SAPS II and APACHE II scores were 48.5 ±11.2 and 27.4 ± 6.3 points, and median LOD score was 7 (5-8) points. Except for Mehta score (p = 0.001), calibration was appropriate in all models. However, discrimination was uniformly unsatisfactory; AUROC ranged from 0.60 for ODIN to 0.72 for SAPS II and Mehta scores. In addition, except for Mehta, all models tended to underestimate hospital mortality. CONCLUSIONS: Organ dysfunction, general and renal-specific severity-of-illness scores were inaccurate in predicting outcome in ICU patients in need for RRT.

Effects of early changes in organ dysfunctions on the outcomes of critically ill patients in need of renal replacement therapy

INTRODUCTION: Acute kidney injury usually develops in critically ill patients in the context of multiple organ dysfunctions. OBJECTIVE: To evaluate the effect of changes in associated organ dysfunctions over the first three days of renal replacement therapy on the outcomes of patients with acute kidney injury. METHODS: Over a 19-month period, we evaluated 260 patients admitted to the intensive care units of three tertiary-care hospitals who required renal replacement therapy for > 48 h. Organ dysfunctions were evaluated by SOFA score (excluding renal points) on the first (D1) and third (D3) days of renal replacement therapy. Absolute (A-SOFA) and relative (D-SOFA) changes in SOFA scores were also calculated. RESULTS: Hospital mortality rate was 75 percent. Organ dysfunctions worsened (A-SOFA>0) in 53 percent, remained unchanged (A-SOFA=0) in 17 percent and improved (A-SOFA<0) in 30 percent of patients; and mortality was lower in the last group (80 percent vs. 84 percent vs. 61 percent, p=0.003). SOFA on D1 (p<0.001), SOFA on D3 (p<0.001), A-SOFA (p=0.019) and D-SOFA (p=0.016) were higher in non-survivors. However, neither A-SOFA nor D-SOFA discriminated survivors from non-survivors on an individual basis. Adjusting for other covariates (including SOFA on D1), A-SOFA and D-SOFA were associated with increased mortality, and patients in whom SOFA scores worsened or remained unchanged had poorer outcomes. CONCLUSIONS: In addition to baseline values, early changes in SOFA score after the start of renal replacement therapy were associated with hospital mortality. However, no prognostic score should be used as the only parameter to predict individual outcomes.

RIFLE classification in patients with acute kidney injury in need of renal replacement therapy.

OBJECTIVES: To evaluate the association of RIFLE classification with the outcomes of critically ill patients with acute kidney injury (AKI) who require renal replacement therapy (RRT). DESIGN AND SETTING: Prospective cohort study in the medical-surgical ICUs at three tertiary hospitals. PATIENTS: 214 patients over 1 year (mean age 71.4+/-15.8 years). Continuous RRT was used in 179 (84%); patients were classified as risk (25%), injury (27%), or failure (48%). Overall mortality was 76%. MEASUREMENTS AND RESULTS: There were no significant differences according to RIFLE classification (risk 72%, injury 79%, failure 76%). Variables selected in multivariate analysis were: older age (OR 1.03, 95% CI 1.01-1.06), presence of comorbidity (3.15, 1.10-9.02), poor chronic health status (6.51, 1.95-21.71), number of associated organ dysfunctions (patients with one or two organ dysfunctions 5.93, 2.03-17.33; patients with three or more organ dysfunctions 26.76, 6.28-114.11), and start of RRT after the first day of ICU (2.46, 1.09-5.53). RIFLE classification was forced into the model and was not selected. However, a subgroup analysis of 150 patients who received mechanical ventilation and vasopressors found failure to be associated with increased mortality (3.58, 1.08-11.80). CONCLUSIONS: Older age, number of organ dysfunctions, presence of comorbidities, and reduced functional capacity were the main prognostic factors. Patients who required RRT after the first day of ICU had a worse outcome. The RIFLE classification did not discriminate the prognosis in patients with AKI in need for RRT.

Comparing strategies to prevent extracorporeal circuit coagulation in continuous renal replacement therapies: role of citrate anticoagulation / Estratégias de prevenção de coagulação do circuito extra-corpóreo em procedimentos contínuos de suporte renal: papel da anticoagulação regional com citrato

Introduction: One important limitation of CRRT is the need of anticoagulation. The aim of this study was to compare the effect of three strategies to avoidcircuit coagulation in CRRT. Methods: Three strategies were analyzed: saline flushes, systemic anticoagulation with enoxaparin and regionalanticoagulation with citrate. Results: A total of 263 membranes used in 72 patients were evaluated. There were 61 (23%) membranes in the citrate group,23 (9%) in the enoxaparin group and 179 (68%) in the saline group. Median lifetime of membranes in the citrate group [48 (27-78) hours] was significantly higher than both of those in the enoxaparin [33 (23-48) hours, p=0.016] and in the saline group [30 (21-61) hours, p=0.008]. There were no difference between lifetimes of membranes in the enoxaparin and saline groups (p=0.604). With the exception of two patients presenting with mild metabolic alkalosis, there were no complication related to the use of citrate. Conclusions: Regional citrate anticoagulation was associated with an increased circuit lifetime incomparison with enoxaparin and saline flushes. With appropriate metabolic monitoring, this method is safe and may become a standard strategy ofanticoagulation in CRRT, even in patients with lower risks of hemorrhagic complications.

Introdução/Objetivos: A necessidade de anticoagulação do circuito extra-corpóreo é uma limitação importante à utilização de procedimentos contínuosde suporte renal (PCSR). O objetivo do presente estudo foi comparar o efeito de três estratégias de prevenção de coagulação em PCSR. Métodos: Trêsestratégias foram analisadas: lavagem com solução salina 0.9%, anticoagulação sistêmica com enoxaparina e anticoagulação regional com citrato. Resultados: Um total de 263 membranas utilizadas em 72 pacientes foi avaliado: 61 membranas (23%) no grupo citrato, 23 (9%) no grupo enoxaparina e 179 (68%) no grupo salina. A meia vida das membranas no grupo citrato [48 (27-78) horas] foi significantemente maior que a do grupo enoxaparina [33(23-48) hours, p=0.016] e salina 0,9% [30 (21-61) hours, p=0.008]. Não observamos diferença significante entre as meias-vidas das membranas no grupoenoxaparina e salina 0.9% (p=0.604). Com a exceção de dois pacientes que apresentaram discreta alcalose metabólica, não observamos complicações relacionadas ao uso do citrato. Conclusões: A anticoagulação regional com citrato esteve associada a uma maior meia vida do circuito extra-corpóreo quando comparada à anticoagulação com enoxaparina e ao uso de lavagem com solução salina 0.9%. Esse método de anticoagulação em PCSR é seguro e pode se tornar a opção de escolha, mesmo em pacientes com baixo risco de complicações hemorrágicas.